Abstract: In recent years, the frequent occurrence of urban waterlogging disasters caused by extreme rainfall due to climate change has posed a threat to urban water safety and sustainable development in China. Accurately grasping the public opinion and emotions in the disaster-stricken areas is of great importance for improving the situational awareness capabilities of emergency management departments in dealing with waterlogging disasters. In today′s era of intelligent networks, the increasing importance of social media as a platform for people to voice their problems and suggestions has made it a major carrier of public sentiment and societal opinion, providing a new avenue for obtaining information about natural disasters. A key technical challenge that needs to be urgently addressed is how to quickly extract urban flood disaster information from social media, and how to perform thematic categorization and sentiment analysis of natural disaster information to accurately grasp the thematic categories of regional disaster situations and public opinion trends. Taking Sina Weibo as an example, this article elaborates on the methods of collecting and pre-processing flood disaster data, and constructs a thematic classification and sentiment analysis model of urban flood disaster information based on FastText to accurately capture the thematic categories and public opinion orientations of disaster-stricken areas. The research results, using the “7.20” heavy rain and flood disaster in Zhengzhou in 2021 as an example show that the methods proposed in this article achieve intelligent extraction and analysis of urban flood disaster data on social media. The theme classification model achieves an F1 score of over 0.80 for the classification prediction of the eight predefined categories, and the sentiment analysis model is generally able to accurately predict data labelled as “negative” in sentiment, which indicates that the FastText-based urban flood disaster information theme classification and sentiment analysis model constructed in this article can meet the needs of urban emergency management departments to dynamically grasp the development of flood disasters and public emotions. It holds significant guiding importance for flood prevention and disaster mitigation planning, calming public emotions, and pinpointing rescue efforts in real time.